Non-stimulus-evoked activity as a measure of neural noise in the frequency-following response

Background: The frequency-following response, or FFR, is a neurophysiologic response that captures distinct aspects of sound processing. Like all evoked responses, FFR is susceptible to electric and myogenic noise contamination during collection. Click-evoked auditory brainstem response collection standards have been adopted for FFR collection, however, whether these standards sufficiently limit FFR noise contamination is unknown. Thus, a critical question remains: to what extent do distinct FFR components reflect noise contamination? This is especially relevant for prestimulus amplitude (i.e., activity preceding the evoked response), as this measure has been used to index both noise contamination and neural noise. New method: We performed two experiments. First, using >1000 young-adult FFRs, we ran regressions to determine the variance explained by myogenic and electrical noise, as indexed by artifact rejection count and electrode impedance, on each FFR component. Second, we reanalyzed prestimulus amplitude differences attributed to athletic experience and socioeconomic status, adding covariates of artifact rejection and impedance. Results: We found that non-neural noise marginally contributed to FFR components and could not explain group differences on prestimulus amplitude. Comparison with existing method: Prestimulus amplitude has been considered a measure of non-neural noise contamination. However, non-neural noise was not the sole contributor to variance in this measure and did not explain group differences. Conclusions: Results from the two experiments suggest that the effects of non-neural noise on FFR components are minimal and do not obscure individual differences in the FFR and that prestimulus amplitude indexes neural noise.